We compared the tolerance-inducing effects of nitroglycerin (NTG) and nicorandil (NIC) in porcine isolated coronary arteries and assessed the role of KATP channels in the response to NIC in nitrate-tolerant and nontolerant preparations. In coronary arteries contracted with U46619 (1-3 x 10(-9) M), NTG, NIC, sodium nitroprusside (SNP), and cromakalim produced concentration-dependent relaxations. The rank order of potency was NTG > or = SNP > cromakalim > nicorandil. Exposure of the rings to NTG (10(-4) M) for 90 min, followed by repeated rinsing for 1 h, produced a parallel, rightward shift of the subsequent concentration-response curves to NTG and SNP; a slight but significant reduction in the maximal response to NTG was also observed. Previous exposure to NTG had no effect on the NIC or cromakalim concentration-response curves. When the tissues were exposed to NIC (3 x 10(-4) M) for 90 min, followed by repeated rinsing for 1 h, there was no effect on the subsequent concentration-response curves to NTG, NIC, SNP, or cromakalim. In both nitrate-tolerant and nontolerant coronary arteries, glibenclamide (GLI 10(-6) M), a selective KATP channel blocker, caused a parallel rightward shift in the concentration-response curve to cromakalim, but had no effect on responses to NTG or SNP. In nontolerant coronary arteries, GLI had no effect on NIC-induced relaxation, but in nitrate-tolerant preparations, GLI produced a significant rightward shift in the NIC concentration-response curve. The results demonstrate that prolonged exposure to NTG, but not NIC, causes tolerance in isolated porcine coronary arteries and that the response to NIC is not affected by nitrate tolerance. The data also suggest that NIC-induced relaxation of nitratetolerant, but not nontolerant, coronary arteries is mediated by activation of KATP channels.